Massachusetts Institute of Technology's Institute for Medical Engineering & Science (IMES) has an immediate opening for a Postdoctoral Associate in Machine Learning for Medicine. We are seeking a highly-motivated individual to join a multidisciplinary research team to develop machine learning techniques for medicine, physiological dynamics modeling, and treatment decision support. The project offers opportunities to develop advanced machine learning, deep learning methods to derive actionable insights from heterogeneous observational data from electronic health records, including clinical time series, medication/procedures, physician notes and reports, and physiological signals.
The ideal candidate will have demonstrated an outstanding capability for independent research and a solid publication record in top-tier machine learning, AI conferences. Candidate must hold a Ph.D. degree in Computer Science, Machine Learning, Statistics, or a related field. Knowledge and experience in one or more of the following areas would be desirable, but not required: deep learning, interpretable models, representation learning, generative models, multivariate time series models, transfer learning, and reinforcement learning. Experience in state-space models, dynamical systems, and physiological signal analysis would be a plus.
Duties will include conducting original research, data assembly and analysis, assisting with writing technical manuscripts, publishing in the peer-reviewed scientific literature, train/mentor students, and collaborating on the development of research proposals.
Successful candidate will work with researchers and faculty members from MIT IMES, CSAIL, and members of the MIT-IBM Watson AI Lab. In addition to a curriculum vitae, applicants should submit a short statement of research interest to Li Lehman, lilehman<at>mit.edu
Li-wei Lehman, Ph.D. Research Scientist Laboratory for Computational Physiology Institute for Medical Engineering & Science Massachusetts Institute of Technology http://web.mit.edu/lilehman/www/